Philips’ New Artificial Intelligence-Enabled Medical Imaging Application

Last week, Philips announced that it received FDA 510(k) clearance for its magnetic resonance for calculating attenuation (MRCAT) imaging and simulation applications. The company’s MRCAT Head and Neck radiotherapy tool will allow radiologists to use MRI as the primary imaging modality for the radiotherapy treatment of soft tissue tumors in the head, neck, brain, pelvis, and prostate.

The advanced Philips system creates images with computerized tomography (CT) density without needing a CT scan, eliminating the time-consuming process of CT-MRI image registration. Pictures generated using the system are also analyzed with artificial intelligence (AI) to create synthetic CT images that enhance the density of the original image for radiologist review.

The accuracy granted with the new system is essential for radiotherapy because mistargeted radiation can damage healthy cells, which is especially detrimental in hard-to-treat cancers of the head and neck.

The MRCAT Head and Neck coil will be used in conjunction with MacroMedics’ imaging mask to combine the new imaging capabilities with the positional accuracy of MacroMedics’ double shell positioning system.

“The superior soft tissue imaging of MR, together with advances in the integration and orchestration of data, including the use of artificial intelligence, promise greater clarity and less subjectivity in planning radiotherapy for head and neck cancer,” said Ilya Gipp, Chief Medical Officer Oncology Solutions at Philips. “Our collaboration with MacroMedics to develop a patient-friendly mask system compatible with our high-resolution dStream imaging coils highlights Philips’ commitment to providing the precision tools needed for the localization and characterization of difficult-to-treat tumors.”

Philips’s plan to integrate MRI technology into clinical radiotherapy planning is a relatively new endeavor. MRIs offer better soft tissue contrast, organ motion visualization, and monitoring of tumor changes than CT scans alone. Using MRI, Philips aims to enhance radiotherapy accuracy, which could help improve patient outcomes.

AI and medical imaging are becoming increasingly intertwined as capabilities improve and allow for better radiology practices. Using AI, researchers from Harvard Medical School and Stanford University were able to detect pathologies as well as human radiologists. That study was significant because it did not require extensive, laborious annotations to achieve high performance.

Recently, Google released its own suite of AI-powered medical imaging tools that have already been implemented at hospitals and health systems. New-Jersey-based Hackensack Meridian Health is using the new AI capabilities to create its algorithm for predicting prostate cancer using medical images. And Hologic, a medical technology company, is working with the Google team to refine an algorithm that identifies precancerous legions.